Alzheimer's disease (AD) represents one of the most important and challenging public health concerns that will affect this country in the coming decades. The ability to identify older adults who are in a preclinical phase of AD will have far-reaching implications for both improved early diagnosis and use of cognitive enhancing pharmacologic therapies. Such treatments aimed at slowing progression ("neuroprotection") will be most effective if applied at the earliest stages before significant neuronal losses have occurred. However, one of the remaining obstacles to such efforts centers on the difficulty in accurately identifying individuals with incipient AD before their deficits are clinically apparent (i.e., prior to both cognitive and functional decline). We propose to recruit and follow a group of nondemented older adults at high risk for AD in an effort to better detect and characterize its preclinical stage. Individuals with mild cognitive impairment (MCI), and individuals with the epsilon-4 allele of the apolipoprotein E gene (APOE epsilon4), both have significantly elevated risks of developing AD. However, MCI and APOE epsilon4 genotype alone are insufficient in determining who will develop AD. We propose to conduct a five-year longitudinal study of MCI- and APOE-related risk in which the combination of initial cross-sectional functional magnetic resonance imaging (fMRI) and genetic assessments will be combined with longitudinal clinical and neuropsychologic outcomes in an effort to identify the most salient predictors of AD. Based on our preliminary studies, we expect that at-risk older adults will exhibit a greater degree of compensatory brain activity during episodic memory encoding than will low-risk normal older adults, and the degree of compensatory activity will be predictive of conversion to AD. In a 2 (MCI vs. normal control [NC]) x 2 (APOE epsilon4 vs. non-epsilon4) design, we will examine blood oxygen level dependent (BOLD) brain activation associated with verbal episodic learning and recall in a variety of brain regions (e.g., medial temporal, frontal, parietal, posterior cingulate). Specifically, the proposed study will (a) examine the relationship between MCI, APOE genotype, and brain signal intensities during episodic memory encoding in the four groups (i.e., MCI/epsilon4; MCI/non-epsilon4; NC/epsilon4; NC/non-epsilon4); (b) determine whether a differential BOLD response in APOE epsilon4 persons--if present--represents a predictor of AD, or is it present across the life span and therefore a normal phenotypic variant independent of underlying AD pathogenesis? This latter aim will be explored by the examination of a cross-sectional young adult sample of APOE epsilon4 and non-epsilon4 persons for comparison of BOLD response patterns to the older adults; and (c) examine for interactions between episodic encoding-related brain activity and other important clinical (e.g., rate of episodic memory decline; conversion to AD), neuropsychologic (e.g., clinical memory measures), and imaging indices (e.g., structural MRI-based medial temporal lobe volumes; arterial spin labeling indices of blood perfusion).